The long-term objective of this research plan is to further our understanding of the neural basis of working memory (WM). Specifically, the goals of this proposal are to characterize the network of functionally connected brain regions mediating visual WM maintenance and determine the behavioral significance of their interactions. Many theories of visual WM postulate the prefrontal cortex (PFC) provides top-down signals to posterior visual association areas in order to keep behaviorally relevant perceptual information activated when it is no longer accessible in the environment. However, evidence for the existence of this interaction is limited and indirect. The experiments outlined in this proposal will use event-related functional magnetic resonance imaging to measure activity levels throughout the brain as subjects perform WM tasks. A recently developed multivariate analysis procedure will then be applied to the data to examine inter-regional interactions during the component stages of WM. The first experiment will seek to determine whether the strength and nature of these interactions is related to behavioral performance. The second and third experiments will assess the mechanisms by which the WM system can dynamically adapt its functional connectivity profile to accommodate increased memory load or overcome distraction.